GPT Image 1 Mini OpenAI vs Vidu Q2 ShengShu Technology

GPT Image 1 Mini

• + Excellent soft lighting consistent with the 'window light' prompt.
• + Clean, minimalist composition with realistic textures on the book and table.
• + Accurate glass refraction and transparency.

• − The plant is more to the side than strictly 'behind' the cube as requested.
• − The glass cube edges are very thin, looking more like a frame than a solid glass object.

Vidu Q2

• + Successfully positions the plant directly behind the cube so it is visible through the glass.
• + Brilliant handling of light and shadows, including the reflection of the blue sphere.
• + The glass cube has realistic thickness and weight.

• − The lighting is quite harsh and 'hard' rather than the 'soft window light' requested.
• − Slightly more cluttered composition compared to Model A.

GPT Image 1 Mini

• + Successfully integrated both source images with high fidelity to the original subjects.
• + Excellent preservation of the man's clothing, hairstyle, and facial features.
• + Realistic lighting and shadow integration within the car's cockpit.

• − The scale of the man relative to the car is slightly too large, making the car look small.

Vidu Q2

• + Dynamic composition and angle that showcases the full car and the coastline effectively.
• + Good motion blur on the road adds a sense of speed.

• − Poor preservation of the man's specific identity and clothing from the source image.
• − Visible AI artifacts on the man's hands and the steering wheel area.
• − The man appears miniaturized within the car's cabin.

GPT Image 1 Mini

• + Excellent adherence to the 'imperfect framing' and '50mm' look with a tight, cinematic crop.
• + Highly realistic skin textures and natural-looking hands.
• + Effective use of bokeh and muted colors to create a moody atmosphere.

Vidu Q2

• + Features more prominent characterful reflections on the wet pavement.
• + Captures the aged texture of the bicycle frame well.
• + Good use of motion blur in the background car.

• − Anatomical errors in the hands and arms, with fingers appearing distorted.
• − The composition is cluttered and the subject's face is cut off awkwardly.
• − Over-saturated colors and harsh lighting reduce the requested cinematic realism.

GPT Image 1 Mini

• + Excellent depiction of battle-worn texture with realistic dirt and weathering
• + Ornate engravings on the armor are intricate and historically plausible
• + Superior lighting and color palette that feels cinematic and warm

• − Hair lacks the specific small beads requested in the prompt
• − Shallow depth of field is a bit too blurry in some foreground areas of the armor

Vidu Q2

• + Successfully includes the requested beads in the hair braids
• + Very sharp and detailed texture on the leather straps and underlayer fabric
• + Good armor reflections with clear light sources

• − The character looks slightly too young and clean for a 'battle-worn' description
• − The armor design feels somewhat more like a digital painting than a lifelike photo
• − The composition is a bit more generic than the atmospheric Model A

GPT Image 1 Mini

• + Excellent text legibility and font choice
• + Clean, perfectly organized grid of high-quality food photography
• + Followed the specific header requests exactly

• − Design is a bit too sparse, lacking descriptive text or prices
• − Empty space on the left feels unfinished

Vidu Q2

• + Contains representative pricing and item descriptions
• + Good use of vibrant accent colors and energetic layout
• + Fits more information into a single view

• − Text is largely gibberish or misspelled
• − Layout is cluttered and lacks white space
• − The food images are repetitive and don't match the specific headers like pizza

GPT Image 1 Mini

• + Excellent typography with a consistent fiery glow effect.
• + Clean and professional layout that feels like a polished advertisement.
• + Accurate currency symbol and starburst design.

• − The burger feels somewhat static rather than dynamic or 'exploded'.
• − The background is less 'fiery' and more of a subtle ember texture.

Vidu Q2

• + Strong sense of motion with sauce droplets and high-intensity fire.
• + The 'exploded' burger effect is more dynamic with several layers and scattering ingredients.
• + Vibrant colors and high visual energy.

• − The currency symbol is incorrect, showing a pound-like hybrid instead of a Euro symbol.
• − The text lacks the requested 'fiery, glowing' effect compared to Model A.
• − The 'LIMITED TIME ONLY' text is slightly cramped near the top.

GPT Image 1 Mini

• + Excellent text legibility and spelling accuracy for all menu items.
• + Consistent chalk texture across all characters.
• + Clean and professional composition that fits the 'handwritten-style' prompt perfectly.

• − The 'elegant cursive' requested for the title is rendered as all-caps print instead.
• − The distribution of letters is perhaps too uniform, making it look slightly like a font rather than organic handwriting.

Vidu Q2

• + Successfully attempted cursive elements for the title as requested.
• + Realistic chalk smudge effects and varied stroke weights add to the 'cozy café' aesthetic.

• − Extreme spelling errors and garbled text across every menu item.
• − Incorrect price for the first item ($34 instead of $24).
• − Visual clutter and overlapping characters make the bottom section unreadable.

GPT Image 1 Mini

• + Successfully captures the facial features and accessories of the character in Image 2.
• + Matches the yellow studio background and red ottoman from Image 1.
• + Correctly adapts the clothing style to the character while keeping the color scheme.

• − Fails to replicate the specific crossed-leg pose from Image 1, settling for a generic crouching pose.
• − Anatomical issues with the foot on the ottoman which appears strangely merged or distorted.
• − The hair texture and style are slightly altered from the source character.

Vidu Q2

• + Excellent adherence to the complex, specific body pose and crossed-leg position from Image 1.
• + High degree of character preservation for the face, sunglasses, and the patterned scarf.
• + Great natural integration of the character's clothing into the dynamic pose.

• − The hand on the left has an extra finger/anatomical distortion.
• − Minor rendering artifacts around the lower legs where they intersect.

GPT Image 1 Mini

• + Excellent cinematic lighting and atmosphere
• + Realistic textures on the spacesuit and horse fur
• + Sophisticated and balanced composition

• − Failed the negative constraint; the astronaut is riding the horse instead of the horse on top

Vidu Q2

• + Vibrant, colorful surreal aesthetic
• + Intricate details in the galaxy patterns on the horse's body

• − Failed the negative constraint; the astronaut is riding the horse
• − The astronaut's hands and the reins are messy and poorly rendered

GPT Image 1 Mini

• + Excellent integration of clothing textures and lighting.
• + Preserves the overall composition and color palette of the beach background.
• + Captures the essence of the vitiligo pattern across multiple limbs.

• − Failed to keep the person's exact face and hair, changing the facial structure significantly.
• − The scarf pattern is simplified compared to the source image.
• − Omitted the sunglasses from the source outfit.

Vidu Q2

• + Successfully included the sunglasses and watch from the reference outfit.
• + Preserved the specific vitiligo pattern on the forehead more accurately than Model A.
• + Attempted to maintain the sand textures on the skin.

• − Fused the coat sleeve into the arm, creating a major anatomical artifact.
• − Significantly altered the facial features and added a mustache appearing to mix both source people.
• − The background beach detail is heavily modified and less realistic than the original.

GPT Image 1 Mini

• + Excellent photorealistic texture on the capybara's fur
• + Cinematic lighting that accurately reflects a night scene
• + Captures the professional, calm expression perfectly

• − Only one paw is clearly visible on the steering wheel instead of both front paws
• − Overall composition is quite dark, obscuring some interior details

Vidu Q2

• + Successfully shows both front paws on the steering wheel
• + Brighter lighting allows for more visible Manhattan street details in the background
• + Great composition showing more of the car's interior architecture

• − The capybara's head looks somewhat pasted onto the body
• − A slight 'uncanny' quality to the businesswoman's facial features

GPT Image 1 Mini

• + Excellent text rendering with no spelling errors.
• + Perfect atmospheric and cinematic lighting that matches the 'moody night sky' prompt.
• + Strong layout that feels like a polished, professional invitation.

• − The 'parchment' texture is very subtle and looks more like a modern poster grain.
• − The twisted trees are somewhat obscured by the dark shadows.

Vidu Q2

• + Captures the 'torn parchment' aesthetic much more literally and creatively.
• + The border elements like webs and thorns are very high contrast and detailed.

• − Contains multiple significant spelling errors in all text fields.
• − Failed to include the correct year (2025 instead of 2026).
• − The composition feels a bit cluttered with chaotic elements.

GPT Image 1 Mini

• + Excellent full head of hair that perfectly matches the instruction.
• + Natural-looking texture and volume.
• + Maintains the overall aesthetic and color palette of the scene.

• − Significantly alters the person's facial features, making him look noticeably younger or like a different person.
• − The hairline and forehead integration looks slightly smoothed over compared to the rugged original skin texture.

Vidu Q2

• + Superb preservation of the original facial features, wrinkles, and skin texture.
• + Successfully adds hair while keeping the identity of the person perfectly intact.
• + The hair texture and lighting match the source image flawlessly.

• − The hairline on the right side (near the forehead) has a minor digital blending artifact.
• − The volume of the hair is slightly less 'full' than Model A, though more realistic.

GPT Image 1 Mini

• + Excellent typography with clean, bold characters.
• + High-quality 3D materials that perfectly match the 'soft refined textures' and PBR request.
• + Superior isometric composition and lighting.

• − The sushi models are slightly simplistic compared to the complex request.

Vidu Q2

• + Successfully included a 3D flag icon as requested.
• + Good variety of sushi types on the plate.

• − The text has minor kerning and alignment issues.
• − The lighting is a bit harsh, creating a more plastic look than 'soft refined textures'.
• − The diorama base has odd small artifacts/blobs at the corners.

GPT Image 1 Mini

• + Features a classic colored-pencil caricature art style with highly exaggerated facial features.
• + Strong source preservation of the subject's denim outfit and general hair structure.
• + Clearly incorporates all requested elements: news anchor desk, microphone, dog, hockey stick, and puck.

• − The transition between the arm and the hand holding the dog is a bit muddy.
• − The hockey stick placement felt slightly like an afterthought at the edge of the frame.

Vidu Q2

• + Excellent integration of the hockey theme by placing the news desk inside a rink.
• + High-quality vector art style with vibrant colors and clear, readable text.
• + Includes multiple dogs and clever details like paw prints on the script paper.

• − Anatomical issues with the hands, including an extra finger on the right hand.
• − The microphone is being held in a slightly awkward, floating manner.

GPT Image 1 Mini

• + Excellent anatomical accuracy for all four animals.
• + Beautiful lighting with soft god rays and realistic fur texture.
• + Clean composition with a focused, cinematic feel.

• − Simple background with fewer flowers compared to the other model.

Vidu Q2

• + Lush, busy environment with many flowers and butterflies.
• + Captures a high energy, playful scene.

• − Anatomical errors including a dog with five legs and a kitten with a rabbit-like tail.
• − Lower level of realism with a more 'digital painting' look than photographic.
• − The fox kit has inconsistently colored limbs and a slightly distorted face.

GPT Image 1 Mini

• + Excellent adherence to the 'hand-painted textures' and 'warm mood' instructions
• + High degree of source preservation with character poses and backgrounds
• + Beautiful soft pastel color palette

• − The colored pencil texture is slightly more reminiscent of children's book illustration than cinematic Ghibli anime style

Vidu Q2

• + Perfectly captures the Studio Ghibli cel-shaded character style
• + Maintains clear lines while adding subtle watercolor washes to the background
• + Excellent preservation of the original image's composition and details

• − The lighting is a bit flat compared to the 'gentle lighting' requested

GPT Image 1 Mini

• + Successfully added wind-blown hair and flying leaves.
• + Maintained the general layout and composition of the source image.

• − Significantly altered the facial features of the woman, losing her likeness.
• − The dog's face and fur texture became less sharp and slightly distorted compared to the original.

Vidu Q2

• + Excellent source preservation, keeping the woman's face and the dog's appearance nearly identical to the original.
• + Convincing hair motion that integrates well with the original hairstyle.
• + Applied vibrant, colorful leaves that add to the energetic feel.

• − One or two leaves have slight transparency artifacts where they overlap the dog.

GPT Image 1 Mini

• + Excellent text rendering with accurate spelling and accents.
• + High-quality vector emblem style with clean lines.
• + Perfect adherence to the banner and established date request.

• − Ignored the request for a light background, providing solid black instead.
• − Minimalist aesthetic is slightly busy due to the hatch-style shading.

Vidu Q2

• + Followed instructions for a light background with subtle texture.
• + Good color palette using the requested warm brown and cream tones.
• + Elegant cloche dome illustration.

• − Severe spelling errors in the restaurant name.
• − Repetitive and redundant text with multiple versions of the name and date.
• − Poor text rendering on the bottom line with distorted characters.

GPT Image 1 Mini

• + Excellent text rendering with no spelling errors.
• + Strict adherence to the requested 6-step logical flow.
• + Clean, professional flat-vector aesthetic with a consistent NASA-inspired palette.

• − The 'Translunar' icon is a bit abstract and messy compared to the others.
• − The layout is slightly unbalanced with a large gap on the middle right.

Vidu Q2

• + Visual style includes nice subtle gradients and depth.
• + Includes consistent iconography across all depicted steps.

• − None of the text is legible or spelled correctly.
• − The numbering system is incoherent and does not follow the requested 1-6 sequence.
• − Displays five astronauts instead of the three associated with the Apollo 11 mission.